The Key to Precision Medicine: Imaging

In his State of the Union address earlier this year, President Obama released initial details of the Precision Medicine Initiative. Launched with a $215 million investment in the 2016 Federal budget, the initiative is aimed at “pioneering a new model of patient-powered research that promises to accelerate biomedical discoveries and provide clinicians with new tools, knowledge, and therapies to select which treatments will work best for which patients.”

While the President did not specifically deem imaging one such tool, it is indeed playing an increasingly significant role in precision medicine, according to James H. Thrall, MD, FACR, chair emeritus, department of radiology, Massachusetts General Hospital, Boston, and a member of the Institute of Medicine of the National Academies.

Thrall discussed this role—and offered a more granular definition of precision medicine—on May 19, when he delivered the Moreton Lecture at ACR 2015, the Annual Meeting of the American College of Radiology (ACR), in Washington, D.C. Four years ago, he told the audience, the National Research Council published a white paper that defined precision medicine as “the tailoring of treatment to the individual characteristic of each patient.”

Key to the practice of precision medicine in this context, Thrall said, is identifying small groups of people who, because they have a similar clinical presentation and biology of their disease, are likely to respond to the same treatments and to have similar prognosis. Identifying which subpopulations respond best to a drug therapy or particular follow-up treatment paves the way for far more efficient clinical trials and, consequently, the ability to treat patients in keeping with the precise characteristics of their disease.

No average patient

Traditionally, Thrall stated, precision medicine has been equated with genotyping, but it must go beyond that because of epigenetics (cellular trait variations caused outside of changes to DNA sequence and other catalysts for variation in gene expression). “Without the ability to classify patients into subpopulations, physicians are forced to consider each as ‘the average patient’,” he explained.  “I think we know that today there is no average patient.”

Phenotyping the physical manifestations of disease to establish cohorts and prognosticate responses to treatments and therapies is a cornerstone of precision medicine, he added, noting that imaging-based phenotyping can trump other phenotyping methods—i.e. clinical history/physical examination, laboratory testing, and histopathology/immunopathology.

“Imaging may often be the earliest and more efficient way to assess response to therapy,” Thrall asserted. It provides “information that is unique, that no other laboratory, pathological, or genetic testing method can duplicate,” especially because it supports the segmentation of patients into highly specific prognostic categories.

For example, at least five clinical trials of a scoring system for patients exhibiting intracerebral hemorrhage are now underway. The system is based on the number of spots of contrast brought to light by CT angiograms. Subjects are segmented into various risk categories based on which of five different scores that correlate with levels of bleeding or not bleeding is assigned them given the findings of their CT angiogram exams.

The phenotype connection

That imaging alone enables sub-phenotyping of patients into four categories—no response, partial response, complete response, and disease progression—contributes equally heavily to its role in precision medicine, Thrall said. He cited the example of two patients with lymphoma, both of whom underwent chemotherapy for the disease and were assessed via multiple FDG-PET scans.

One patient initially had a partial response to the treatment but was later found to be experiencing disease progression. In the second patient, FDG-PET showed a complete metabolic response to the chemotherapy regimen.

Traditionally, Thrall observed, the international scoring system served as a prognostic indicator for lymphoma patients, but FDG-PET scans may be more accurate; regardless of a patient’s international prognostic score, a negative follow-up scan signals an “extremely good likelihood” of progression-free survival.

“There is no other method in medicine to provide this kind of information,” Thrall stated. “The pathologist cannot make this prediction, (and) the geneticist cannot make this prediction. We make this prediction by imaging.”

Of particular note, Thrall added, is the fact that imaging-based phenotyping appears to be of great value in researching therapies for Alzheimer’s disease. Here, it can help to assess the efficacy of treatments aimed at halting or decreasing the accumulation of amyloid plaque in the brain.

Clinical trials of about a half-dozen drugs that are intended to stop further amyloid plaque accumulation or reduce the volume of amyloid plaque already found to have accumulated are presently being conducted.  Imaging, in conjunction with clinical psychometric testing, can serve as a determinant of whether these drugs are fulfilling their purpose, according to Thrall.

Fighting cancer

Thrall also sees a major role for imaging on the oncogenetic side of precision medicine. He cited the case of actress Angelina Jolie, who was found to have inherited a BRCA1 gene mutation from her mother. After learning that she had an 87% risk of breast cancer and a 50% risk of ovarian cancer, Jolie elected to undergo a prophylactic bilateral mastectomy as well as a bilateral salpingo-oophorectomy.

Given these odds, the removal of Jolie’s breasts, ovaries, and fallopian tubes was almost definitely the proper course of action, Thrall observed. Conversely, he noted, another woman with the identical mutation, but a much lower risk of developing breast cancer than Jolie along with a strong desire to breastfeed her future children, would not be subjected to surgery under a precision medicine umbrella. Rather, she would be monitored via regularly scheduled imaging exams to track the occurrence, location, extent, and severity of the disease.

Thrall believes that radiologists “have actually been doing the kinds of things inherent in precision medicine forever—probably far more than they may realize. “When a radiologist renders a report, the report is actually a description of the imaging phenotype of disease manifestation, but we’ve just never thought about it that way before,” he concluded. “We are already one of the supreme methods for phenotyping. We are playing an important role. And we are learning more and more about the relationship between imaging phenotype and genotype every day.”